ODD-Based Long-Term Decision-Making for Intelligent Vehicles
DOI: 10.1109/iv64158.2025.11097482
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This paper addresses the challenge of ensuring safe and efficient navigation for Intelligent Vehicles (IVs) in complex, dynamic environments by integrating the Operational Design Domain (ODD) into long-term tactical decision-making. Current approaches often focus on immediate dynamic constraints or react to ODD violations only after they occur, potentially leading to risky situations or loss of autonomy. The authors propose a framework called ODD-aptive, which evaluates vehicle capabilities across both current and future states to construct a reachable horizon. This ensures the vehicle remains within its ODD, preventing imminent departures from safe operational limits due to contextual changes like weather or traffic density. The proposed architecture operates at the tactical level, modeling decision-making as a Markov Decision Process (MDP). The system utilizes a Sense-Think-Act paradigm to assess the vehicle’s decisional capability for specific maneuvers. A key component, the Maneuver Capability Estimator, decomposes maneuvers into skills (e.g., perception, reasoning, actuation) and filters out actions that are not feasible given the current Operational Domain (OD). This filtering significantly reduces the MDP’s search space by excluding unsafe or unreachable states. The MDP includes states defined by the OD and dynamic scenario, actions such as Keep Lane, Lane Change, and Minimal Risk Maneuver (MRM), and a reward function that penalizes collision risk, off-track risk, and states with limited actionable options. The optimal policy is computed online using Value Iteration. The framework was evaluated using simulations in IPG CarMaker. The test scenario involved a highway drive where Vehicle-to-Infrastructure (V2I) communication provided prior knowledge of dense fog 800 meters ahead, which degrades perception capabilities. The simulation demonstrated that the system proactively adjusted its behavior to avoid states where lane changes would be unsafe due to reduced visibility. Specifically, the vehicle executed lane changes earlier to position itself in lanes that allowed for safe stopping or continued operation, avoiding states where the lack of perception skills would restrict available actions and prevent necessary MRMs. The results showed that the state-dependent filtering reduced the number of explored states and computation time compared to standard exploration methods, while ensuring the vehicle remained in safe, operational states despite the environmental degradation. The significance of this work lies in its proactive approach to ODD compliance, shifting from reactive safety measures to long-term planning that anticipates capability limitations. By formalizing the relationship between contextual changes and vehicle skills, the architecture enhances the transparency and reliability of autonomous decision-making. This method allows IVs to maintain functionality and safety in evolving conditions, supporting higher levels of autonomy by ensuring that tactical decisions are always grounded in the vehicle’s actual operational capabilities. The study highlights the importance of considering inter-dependencies between environmental dimensions and system functions to prevent dangerous scenarios before they arise.
Provenance
The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | Crossref | — | — | 1 | 2026-06-25 |
| archive | success | unpaywall | — | — | 2 | 2026-06-26 |
| extract | success | cached | — | — | 2 | 2026-06-26 |
| clean | success | clean | — | — | 1 | 2026-06-26 |
| chunk | success | chunk | — | — | 1 | 2026-06-26 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-26 |
| enrich | success | openalex | — | — | 1 | 2026-06-26 |
| promote | success | — | — | — | 1 | 2026-06-25 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 1 | 2026-06-26 |
| tag | success | vector_similarity | — | — | 6 | 2026-06-26 |
| verify | success | — | — | — | 1 | 2026-06-26 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
Information type
What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).
- Theoretical Contribution: computational model